Four real AI use cases in Australian manufacturing and the hardware decisions behind moving from pilot to production. 

In this episode, Troy Smithells from Xenon Systems (Australia's only Nvidia Elite partner) walks through where AI is actually earning its keep in Aussie manufacturing, and why production workloads often need to leave the cloud. We cover: 

Hand-drawn to CAD file conversion: 4x faster, near-100% accuracy, $750K per year saved at a steel fabricator 

Physical AI in pipe inspection: tiny robots mapping up to a kilometre of pipe for leak detection and preventative maintenance 

Predictive maintenance from sound: a listening device on a conveyor belt catching wear before failure 

Computer vision for construction OHS: PPE checks, exclusion zone alerts, and yes, crocodile alerts on the docks at Cairns 

Why cloud AI costs scale exponentially and on-prem or hosted GPU hardware can land at roughly a quarter to half the spend 

Data sovereignty, the US Cloud Act, and the space, power and cooling reality of running GPU clusters (132kg, 110dB, and a burn-in you need earplugs for) 

When to bring in infrastructure specialists: after the proof of concept, before you try to scale 

Whether you're a plant manager, operations leader, CTO, or integrator weighing up an AI project, this conversation gives you a practical map from pilot to production grade AI. 

Follow Manufacturing Tech Australia on Spotify, Apple Podcasts, or YouTube, and leave a review if this episode helped you. 

Show notes and links for this episode: https://www.manufacturingtech.au/episode/79-from-cloud-to-concrete-practical-ai-use-cases-in-manufacturing-with-troy-smithells/